Hello,

I'm doing LES for a backward-facing step. I want to impose periodic boundary condition in spanwise direction (on both side walls). However, it seems I can't apply periodic boundary condition.

I can apply wall condition or symmetric condition. But I don't think it is a good boundary condition for LES.

Would anyone help me?

Thanks in advance.

Sungho

Why is this not a good BC for LES?

I think its very good BC, it will reduce your computation time. And people often use periodic BC for side walls. If you look through literature you will find for channel flows periodic BC is almost always used for side walls. Can you please explain why you think its not a good idea.

Can you please let me know if you are using any paper to validate your results? Thank you

I want to apply periodic boundary condition and I believe it is a good boundary condition.

What I meant was I can't find periodic boundary condition in CFX for backward-facing step. (I can use only symmetry, wall, etc...)

I'm trying to validate my data with the following two papers.

1. Direct numerical simulation of turbulent flow over a backward-facing step, JFM, 1997, vol 330, pp 349-374 - HUNG LE a1 , PARVIZ MOIN a1 and JOHN KIM

2. Backward-facing step measurement at low Reynolds number, Re h= 5000 S Jovic, DM Driver - NASA Technical Memorandum, 1994 - ntrs.nasa.gov

Hi,

Make sure your mesh faces match on both sides of the periodic interface so you can use a 1to1 interface. Otherwise you have to use a GGI and that is not recommended, especially for LES applications.

Glenn Horrocks

Thank you! I got it.

Sungho Yoon